|
Activity Number:
|
26
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
General Methodology
|
| Abstract - #310405 |
|
Title:
|
A Study of Robust Estimation Approach for Analysis of Variance When the Distribution of Error Terms Is Non-Normal
|
|
Author(s):
|
Aysun Çetinyürek*+ and Bridal Senoglue+
|
|
Companies:
|
Baþkent University and Ankara Universities
|
|
Address:
|
Baglica Kampusu, Ýstatistik ve Bilgisayar Bilimleri Bolum, Ankara, 06530, Turkey dogol cuddesi Fen Faksltesi, Ankara, 06100, Turkey
|
|
Keywords:
|
modified likelihood ; nonnormality ; beta distribution ; robustness
|
|
Abstract:
|
Analysis of Variance (ANOVA) is used to test the equality of the means when there are several groups. In ANOVA procedures, traditionally the error terms are assumed to be normally distributed. However, non-normal distributions are more prevalent in practice. In this paper, we derive the estimators of the model parameters in one-way and two-way ANOVA when the distribution of error terms is Beta. For Beta(a,b) distribution, the maximum likelihood (ML) method does not provide explicit estimators for the model parameters. Explicit estimators are derived via modified maximum likelihood (MML) methodology by linearizing the intractable terms in likelihood equations. MML estimators are known to be asymptotically fully efficient in terms of the minimum variance bounds (MVBs) and they are also robust. Hence, we propose to use MML estimators in ANOVA models when the error distribution is Beta(a,b).
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
